The Hymenoptera are a major order of holometabolous insects. That is, they undergo complete metamorphosis with distinct egg, larval, pupal, and adult stages. They are one of the five megadiverse insect orders along with Coleoptera, Diptera, Lepidoptera, and Heteroptera, and perhaps even the most species rich of any insect order—certainly this is true at temperate latitudes. They are generally cosmopolitan and, except for some specialized groups, they are most speciose in the tropics.
Although after working with Hymenoptera for a while it becomes easy to recognize members of this order, there are almost no conspicuous defining characters, because the majority of hymenopteran attributes are plesiomorphic; that is, they are shared with the common ancestors of various other orders. It is not surprising therefore that although almost everyone on earth, save perhaps those living at extreme northern latitudes, is familiar with ants, bees, and social wasps and has vernacular names for these particular taxa, there is not a single vernacular name in any language that refers to them in toto. Hymenoptera are also diverse in terms of their life histories: they include phytophagous, parasitoid, and predatory taxa, both solitary and highly social species, and they range in size from the rather large and intimidating spider-hunting pompilid wasps that can reach 12 cm wingspan down to the tiniest parasitic wasps that are approximately 0.1 mm in length (males of the wingless mymarid chalcidoid, Dicopomorpha echmepterygis). It is hard to overstate their ecological importance because they are collectively involved in so many types of interaction, and it is likely that many are effectively keystone species in their own habitats.
The Hymenoptera get their name from the Greek words humen and pteron, meaning membrane and wing, respectively, and this gives the first clue to identifying them. Excluding the numerous exceptions of apterous and brachypterous species that are widely distributed through the order, hymenopterans possess two pairs of membranous wings that are devoid of scales. The forewings are larger than the hind wings, and the two are interlocked during flight by a row of special hooks called hamules (or hamuli) that are on the anterior margin of the hind wing; these hamuli engage (or interlock) with a fold on the posterior edge of the forewings. This system makes the Hymenoptera functionally dipterous (two-winged) during flight, since the wing surfaces on either side of the body acts as a single aerofoil. Hamules are unique to this order of insects.
GENERAL BIOLOGY
Since Hymenoptera is a very large order, it is not surprising that a considerable number of biologies and life history strategies are exhibited by its various taxa. Broadly speaking, the basal lineages are phytophagous as larvae, feeding both ecto- and endophytically on a large range of herbs, shrubs, and trees; a few tropical pergid saw-flies even feed on slime molds! The great majority of the remaining species are either parasitoids of other insects or predators of insects (e.g., the yellowjackets or social wasps, which are members of the Vespidae) or spiders. However, among the higher taxa, there have also been several reversals to phytophagy, especially through the formation of galls on plants (cecidogenesis). The bees (Apidae) and one other, small tropical group, the Masarinae within the Vespidae, have evolved to make use of pollen and nectar as a larval food source.
Development
Like other holometabolous groups, hymenopterans primitively pass through generally five instars, though the number of instars is typically smaller in endoparasitic taxa, and in one such genus there seems to be just a single instar. The final instars of hymenopterans are rather morphologically conservative, with most sawflies having rather caterpillar-like larvae with well-developed true legs and variously developed prolegs on several of the abdominal segments. Adoption of an endophytic way of life by cephoid sawflies and wood wasps was accompanied by a reduction in the prolegs and more generally by reduction of sensory structures. Final instar apocritan wasp larvae are all quite similar and are termed hymenopteriform. They are superficially rather maggotlike in that they lack legs and other processes and often have a rather reduced head. However, many endoparasitoids have highly bizarre, first instars characteristic of their particular families, and for which a variety of specific terms have been coined.
The pupal stage of hymenopterans is exarate; that is, the antennae, legs, and wings are free from the body (in contrast to the Lepidoptera, e.g., in which these components are fused with the body). The pupae tend to be rather delicate and are easily damaged. All sawflies and most members of the Ichneumonoidea + Aculeata clade produce a silken cocoon to protect the pupa. Most of the other parasitic taxa do not, however, probably because they pupate within the host remains or, if they pupate externally, they do so in a location where the pupa is likely to be protected by the surroundings, such as within a leaf mine, gall, or wood boring.
Key Features in Hymenoptera Evolution
Given the huge size of the order, it is interesting to consider what features have enabled hymenopterans to be so successful in terms of both individuals and total number of species. Most attention has focused on a small number of features such as selection of oviposi-tion site, modification of that site, the use of venoms, and the evolution of the thin wasp waist, all of which are discussed in this article. In addition, the unusual form of sex determination mechanism, hap-lodiploidy, may have been particularly important in the evolution of sociality. It is likely, however, that few of these traits have operated in isolation, and it is the interactions of these and other factors that have been important. Thus, for example, evolution of sociality may have been facilitated by the sex determination mechanism but also requires the abilities to remember where the nest is, to recognize nestmates, and to be able to defend the nest.
Several studies have emphasized that the success of the Hymenop-tera has probably been a consequence of the general tendency of these insects to provide their offspring with particularly nutritious food sources, and when necessary (and that has been often) to modify poorer foods to better ones. Although this may be most familiar in terms of the provisioning of larvae in the nests by the social wasps and bees, such behaviors and physiological adaptations are to be seen all through the order and are manifested in many different ways.
First, there is egg placement and the larval food resource. The morphology of the ovipositor has been crucial in this respect. The hymenopteran ovipositor is used not only for laying eggs, it is also used to pass venom and/or other secretions to the place of oviposi-tion. In the parasitoid taxa, these venoms either cause paralysis of the host or are important in overcoming the host’s immune response against the parasitoid. The ovipositor is typically well supplied with sensilla, and the insects receive and interpret the resulting sensory information and use it in deciding whether they have located a site or host suitable for egg laying. This organ has been especially well studied in parasitoid taxa, and such observations have been used to test many evolutionary concepts.
In the majority of the aculeates (stinging wasps, bees, and ants) the egg-laying role has been lost, but the same structures are still present and are used for envenomation of prey or enemies. The venoms of most of these act on the nervous systems or nerve-muscle junctions of their prey insects, permanently paralyzing them. In this sense, the venoms are rendering their larval food sources manipula-ble and safe by preventing the prey insect from wriggling or moving to damage the wasp’s developing young.
“Venoms” were important even before the evolution of parasi-toidism. For example, at least some and possibly most wood wasps inject chemicals into their host trees along with their eggs and symbiotic fungi fragments, and these toxins probably either kill the living cambium cells or in some other way help the symbiotic fungi to overcome the trees’ defenses so that the wood wasp larvae can feed on the developing nutritious fungal hyphae. As often happens, these conclusions are based on relatively few data, and observations of other species are very much needed.
Evolution of the thin wasp waist, which defines a large group of families called the Apocrita, was another absolute key feature in that it greatly increased the mobility of the posterior abdomen relative to the thorax. This in turn allowed greater control of the ovipositor and greater variety in its use; later, it allowed the sting, which is in fact just a derived ovipositor, to be much more effective as a weapon of defense and offense. It is interesting that vertebrates can learn that a bee or wasp can deliver a sting and that part of the recognition of this ability involves the very conspicuous abdominal movements of the insect as it probes for a vulnerable spot with its sting. Because male Hymenoptera never possess stings (because the males do not have an ovipositor-derived apparatus!), they are harmless in this respect. Often, however, males very effectively mimic female wasp stinging movements such that people, and probably many experienced predators, do not take the risk and quickly release them—this behavior has been termed aide-memoire mimicry.
The wasp waist, contrary to many people’s initial expectations, is actually not located between the thorax and abdomen, but is a constriction between the first and second abdominal segments (Fig. 1). In the ants, posterior abdominal mobility is increased even more by second and sometimes third constrictions between the second and third, and third and fourth, abdominal segments, which give rise to the distinct node or nodes between the middle and posteromost body regions. There is a very good reason for the wasp waist to be positioned after the first abdominal segment. Higher hymenopterans are typically strong fliers, and their longitudinal flight muscles are consequently large. Because these muscles are attached internally on the anterior of thorax (actually the mesonotum) and posteriorly on a large internalized chitinous phragma that slants posteriorly, if there were a constriction immediately behind the last (third) thoracic segment, the size of the flight muscles would be greatly restricted. By having the first abdominal segment fused to the thorax, larger flight muscles can be accommodated. Thus, the middle body part of an apocritan hymenopteran is comprised of the pro-, meso-, and metathorax, plus the first abdominal segment, the latter being termed the propodeum. Of course, this nomenclature has often led to confusion among those who lack detailed familiarity with wasp physiology. Nowadays, to avoid ambiguity, it is becoming increasingly common to refer to the middle body region as the mesosoma and the part behind it as the metasoma. Further nomenclatural confusion can arise when, as in the ants and some parasitic wasps, the first metasomal segment (i.e., the second abdominal segment) is greatly reduced. The most conspicuous part of the metasoma is then referred to as the gaster.
Venoms
Most if not all Hymenoptera, even the sawflies (which are phytophagous), have a “venom” gland associated with the ovipositor or sting. In fact, most have at least two distinct glands, the venom gland proper, also referred to as the acid gland on account of its typical his-tological staining properties, and a second gland, referred to as the alkaline gland or Dufour’s gland in the Aculeata (see later). Details of the function of the secretions of these glands are known for only a few species; for the sawflies, such knowledge is almost completely
FIGURE 1 Stylized illustration of an apocritan hymenopteran showing the wasp waist between the first and second abdominal segments. The first abdominal segment, called the propodeum, is broadly attached to the thorax (colored blue), and the combined structure is referred to as the mesosoma. The abdomen (darker shading) posterior to the mesosoma is often called the metasoma.
lacking. It may reasonably be assumed that the initial function was production of lubricants that assisted passage of the egg down the ovipositor; but natural selection would have acted quickly to favor organisms that showed an ability to benefit from modification of the substrate.
In the parasitic wasps, the venom gland products are thought to be primarily associated with two roles: to help overcome the host’s immune defense against the parasitoid’s egg or larva, and sometimes, especially among the ectoparasitic idiobionts, to paralyze the host. In some taxa, perhaps in many, there may also be secretory products from parts of the female reproductive tract itself, and these may play important roles in overcoming host immunity. Of great interest among these are the “polydnavirus soups” produced by the calyx gland in a few groups of Ichneumonoidea; this gland is a modified part of the lateral oviduct.
The venom glands in the aculeate wasps are the source of the well known, pain-inducing toxins that many social and some solitary Hymenoptera use to such good effect in self-defense. One solitary aculeate, a mutillid wasp (or velvet ant), is commonly called the camel-killer because its venom is reputedly strong enough to have that effect, and there are anecdotal reports of soldiers who have been incapacitated by the pain caused by encounters with this substance. Typically, however, the venoms are rather less fearsome. These pain-causing venoms are very specialized and contain a variety of neurotoxins, including, in some taxa, small ringlike peptides that insert themselves in cell membranes and cause depolarization of nerve cells, and consequently pain. It is easy to envisage how these peptides could have evolved from toxins that were originally selected to cause paralysis of the arthropod prey of their ancestors, although probably most are so highly modified that any initial similarity has been lost. A few of the larger parasitic wasps have also developed pain-causing venoms for defense, but their stings are quite mild and the effects short-lived compared with those of many aculeates.
The main function of Dufour’s gland seems to be the production of substances that are involved in intraspecific communication. In the parasitic wasps, the gland probably serves primarily as a source of marking pheromones that indicate where an egg has been laid, likely to minimize self- and intraspecific superparasitism. Among the social aculeates, the functions of this gland have been greatly elaborated, and it is the source of many other pheromones that are involved in colony organization.
The Ovipositor: A Key Organ
Since Hymenoptera in general are known to take great care in the placement of their eggs, it is not surprising that the ovipositor is an important organ, and one that has shown many specializations for particular modes of life.
The hymenopteran ovipositor is derived from abdominal appendages and comprises three independently movable parts, called valves, that together form the egg canal. The dorsal valve is a fused structure, but the ventral ones are separate. There is no intrinsic ovipositor musculature; rather, the movements of the valves depend on muscles within the abdomen that pull on the internal apodemes of the three valves. Nevertheless, several parasitic taxa have evolved mechanisms that enable them to steer their ovipositors and thus increase their chances of successfully attacking a mobile host that might otherwise be able to wriggle away from its attacker. Although the penetration of the substrate by wasp ovipositors is usually referred to as “drilling,” it is important to realize that there is no circular motion: penetration is achieved by the to-and-fro motion of the three valves relative to one another. In the simplest mode of operation, one valve has a projection or nodus that interlocks with the substrate, and this acts as a support for the others to be pushed forward.
The sawflies get their name from the laterally compressed, strongly serrated, ovipositors with which they insert their eggs under plant cuticle. These ovipositors are unsuited, however, for penetration of wood, and the wood wasps’ ovipositors are longer and rounder in cross section, with serrations used for rasping wood fibers, located just at the tip (see Fig. 2). Most of the parasitic Hymenoptera have
FIGURE 2 A large ichneumonid wasp, Megarhyssa sp., using its ovipositor to “drill” through a tree trunk to reach its host, a siricid wood wasp larva. The ovipositor is very thin and pointing between the fore legs; the large black structures are the protective ovipositor sheaths.
a similar ovipositor except that in many of those with exposed hosts it is much shorter and has reduced serrations because there is no substrate to “drill” through. Some ovipositors are very long (up to 12 times longer than the wasp’s body), and various mechanisms and behaviors have evolved to enable the wasp to manipulate them.
The Road to Parasitoidism
The evolution of parasitoidism has long been of interest, and several possible scenarios have been discussed at various times. As our understanding of the phylogenetic relationships of the basal parasitic wasps has firmed up over recent years, it now seems most probable that the transition to a parasitoid way of life occurred first among some ancestral wood wasp, because the closest extant sister group of the parasitic Hymenoptera is almost certainly the Orussidae; which collectively are derived from wood wasp ancestors. The most widely discussed and generally accepted proposal for the evolution of parasitism in the Hymenoptera envisages an ancestral wood wasp gaining an advantage by producing a larva that could encounter, kill, and eat another wood-boring insect—possibly the larva of another wood wasp or of a beetle—because such a food item would have a greater nutritional value than the plant diet. An advantage therefore would have been gained if the female ancestral parasitoid were to seek out for oviposition sites branches where such food bonuses occurred. This could occur only if the prey item, such as a batch of eggs or perhaps a minimally mobile prepupa or pupa, did not pose a danger to the ancestral parasitoid. The next step would be from a facultative utilization of vulnerable prey insects to supplement a plant-based diet with an obligate one, eventually eliminating the need to consume plant material. So by evolution of the wood wasps, prey species become hosts. At first, as in extant orussids (see later) the larva might have done the final prey location, but the protoparasitoid in this scenario always evolves to use its ovipositor to injure or kill the prey, perhaps by stabbing. Subsequently, venoms evolved greater sophistication until they were able to induce permanent host paralysis, and even to help keep the host fresher for longer by preventing it from becoming infected with fungi and bacteria.
Gall Formers
As with several other insect orders, some groups of Hymenoptera have evolved ways of making plants produce especially safe and nutritious places (i.e., plant galls) in which to shelter their young. Because hymenopterans produce venoms that contain a wide range of pharmacologically active compounds that affect host insect physiology, it is not very surprising that some of these might affect plants—and the typical plant response to damage is cell proliferation, forming a callus. Thus, venom usage may have preadapted parasitic wasps for the evolution of gall forming (cecidogenesis). We cannot know whether this has always been true, and at least in the extant Cynipidae, the gall wasp family, the phytotoxins that stimulate gall production are produced by the wasp’s larva and are not components of the female venom. Gall-forming sawflies, however, do seem to have undergone this preadaptation. True gall wasps (cynip-ids) have colonized and diversified upon only a few plant groups, notably Rosaceae and Fagaceae, although the other gall-forming Hymenoptera collectively attack a huge range of plants.
One family of chalcidoid wasps, which are also gall formers in a sense, are of particular interest. These are the pollinating fig wasps belonging to the Agaonidae. These insects develop within the ovaries of fig flowers, consuming as larvae the galled tissue. Many people eat figs, but few know that originally at least, the cultivated fig and all its relatives in the genus Ficus (Moraceae), relied absolutely on the pollinating activities of these specialized agaonid wasps. Figs are actually not fruit in the strict sense but are syconia—that is, a hollow flasklike structure containing many flowers. The female fig wasps collect pollen from the flowers within the fig in which they have developed, and upon entering another developing fig, actively pollinate the flowers. Figs and their pollinator wasps have evolved in a tight association, and each of the 400 or so species of fig (all members of the genus Ficus) has one or, rarely, a couple of pollinating wasp species associated with it that pollinate no other fig species. In these wasps, the males are especially highly modified and do not leave the interior of the fig as adults. Instead they compete with one another for mates and frequently kill competitors with their large mandibles. Males are also responsible for chewing an exit hole through the wall of the fig that enables the female pollinator fig wasps to escape and go in search of new figs.
The Road to Sociality
Perhaps the greatest claim to fame of Hymenoptera is that the order includes several highly social groups of insects, various bees, yellowjacket (vespid) wasps, and ants. The only other insect order with such a large number of highly social species is the termites (Isoptera). Within the Hymenoptera, sociality has evolved on quite a few separate occasions, and much consideration has been given to the reasons for this circumstance. Probably a variety of factors have contributed.
Provision of a good food source for the larvae first involved selection of a suitable host plant species, but with the evolution of parasitoidism, the degree of selectivity increased. Further selectivity is apparent with the very narrow host ranges of many of the parasitic taxa and, also, their ability to assess the suitability of individual hosts of the right species. This may be viewed as a progression in the degree of individual attention provided to each individual offspring. In several lineages, all in the Aculeata, additional behaviors have evolved to make hosts, in a sense, more suitable—at first these changes consisted of moving a host to a slightly preferable location before oviposition. This is seen in several members of the Bethylidae. Some bethylids also show a degree of parental care in that the female remains with her single brood through their development to guard them against predators and to guard the host against other parasitoids including conspecific females. In the bethylid wasps the degree of parental care is limited to the brood care, but in a number of groups, notably among the pompilids and sphecids, the female wasp prepares a hideaway in which to cache the insect or spider that will provide her offspring with food. Probably at first, once a host had been identified, the female would locate, dig, or modify a burrow; a further evolutionary step was likely the postponement of burrow construction until the search for a host had begun. This stage required the behavioral sophistication of being able to remember the location of the burrow and to relocate it once a prey had been found. This step was crucial for the evolution of sociality because nest members must be able to locate their nests after foraging expeditions. Another major development allowed by this evolutionary advance was the use of hosts, now usually called prey, that are smaller than what would be necessary for the development of the wasp, because it was now possible to bring back multiple individual hosts for each of the wasp’s larvae. This is the stage exhibited by many sphecid wasps.
In these, the paralyzed prey are first accumulated until there are sufficient, then an egg is deposited on the cache, and this set is sealed into a cell, after which the female starts collecting more prey for her next egg. Bees and vespid wasps have independently dispensed with sealing their larvae in individual closed cells with all the food that they need; instead, they provide food continuously upon demand— prepared food, that is, rather than whole prey individuals.
Reproduction and Sex Determination
As far as is known, all hymenopterans have a haplodiploid sex determination system, which means that haploid individuals (having only one set of chromosomes and resulting from unfertilized eggs) are males, whereas diploid individuals (having two copies of each chromosome and resulting from fertilized eggs) are females. This sex determination system is found in a few other groups of organisms, notably in thrips and some rotifers. Development of unfertilized eggs into males is a form of parthenogenesis and is termed arrhenotoky.
There is certainly more than one sex determination mechanism even within this haplodiploid system, and this has important consequences in matters such as biological control. The best understood (or surmised) mechanism, called complementary sex determination (CSD), is characterized by the occasional occurrence of diploid (but infertile) males and by the tendency of the proportion of these to increase with inbreeding. Culturing species with this sex determination mechanism is difficult because small population sizes result in the gradual loss of sex alleles and so an increase in frequency of diploid males and loss of colony vigor. In the rearing of insects for biological control programs, the appearance of an abnormally high number of diploid males can be a very serious setback, because when colonies become inbred, they go extinct.
COMPLEMENTARY SEX DETERMINATION (CSD)
Although the exact molecular details are unresolved, there is a reasonable hypothesis that CSD may involve polymeric proteins, the most simple of which are dimers. A heterodimer (i.e., an association of two different protein chains) has a different shape from a homodimer (two identical protein chains), and this shape difference determines the sex of the offspring. Because haploid individuals can make only one form of the protein (they only have one gene locus), they must make the homodimeric form, and this means that they will develop as males. It seems that in natural populations there are typically quite a few sex alleles (roughly between 6 and 50), with the result that the proportion of fertilized eggs that contain two copies of the same allele is rather small. Thus the proportion of eggs that either fail to develop or produce infertile diploid males would be expected to be small, as well.
NONCOMPLEMENTARY SEX DETERMINATION
Not all Hymenoptera can possess CSD, as is evidenced by the routine inbreeding that occurs, for example, in some parasitic wasps. In these species, a female lays on or in a single host an often large brood of eggs consisting mostly of daughters (from fertilized eggs) and a single haploid male (or at least a very low number of sons), which fertilizes all his sisters when they emerge. This goes on for many generations and undoubtedly must lead to increased homozygosity, but these wasps show no progressive change in sex ratio or fecundity as CSD would necessarily cause. However, what sex determination system is involved in these insects is not known, and while a gene dosage mechanism is widely postulated and seems highly likely, it is not proven.
RELATEDNESS AND MATING
A particular consequence of haplodiploidy that has been invoked as a major reason for the multiple independent evolutions of sociality within the Hymenoptera is that it leads to a change in the normal degrees of relatedness between a mother and her sons and daughters and between siblings, and this difference is most pronounced if the female has mated only once. The reasons are as follows. A male offspring gets all his genes from his mother (because he is haploid and so has no male parent), but he gets only half of the mother’s chromosomes; in the Hymenoptera, therefore, a son is 50% related to the mother just as in mammals. However, if the female has mated only once, her daughters, which come from fertilized eggs, contain half her set of chromosomes plus 100% of those from the male parent (because he is haploid and all his sperm are identical). Overall, therefore, each daughter is 75% related to each other daughter, whereas daughters are related to their mother only 50%. The argument regarding sociality is that because sisters in Hymenoptera are more closely related to each other than they are to any potential offspring they could have, their fitness will be better enhanced by helping their mother to produce more sisters than by reproducing themselves.
The vast majority of species of social ants, bees, and wasps only mate once. Thus, the foregoing arguments may generally hold. However, there are exceptions, and honey bee queens typically mate about a dozen times. Thus, for these insects the disparity in related-ness between offspring and sisters is much closer to the 50:50 ratio of normal diploid taxa.
PHYLOGENY, CLASSIFICATION, AND WHAT PARTICULAR GROUPS DO
The Hymenoptera are member of the monophyletic group of insects known familiarly as the Holometabola, and it is among these therefore that their relationships must be sought. The search for such relationships, however, has thus far proved to be rather difficult, and despite a growing body of molecular data, a consensus about the exact relationships of the order has yet to emerge. Weak evidence has been put forward to suggest a relationship with the Mecoptera, but the characters involved are liable to homoplasy. One possibility is that the Hymenoptera form a sister group to the whole of the rest of the Holometabola (or at least, of the extant holometabolan orders).
Classification
For a long time the Hymenoptera have been broadly divided into three groups: the sawflies or “Symphyta,” the aculeates; or stinging and social wasps, bees, and ants and their close allies; and the remainder, which are typically referred to as the parasitic wasps or “Parasitica,” even though many of them are not parasitic at all. It has long been realized that this is an unnatural arrangement, and slowly an attitude more consistent with our understanding of hymenopteran phylogeny has been filtering into the literature. Thus, although it is still useful to be able to refer to sawflies as a group, the use of a formal classificatory term for them to the exclusion of all other Hymenoptera (i.e., Symphyta) is rather unsatisfactory. This is because it has long been known that the “Symphyta” are a basal grade that leads to the wasp-waisted Hymenoptera; “Symphyta” are therefore not a mono-phyletic group (i.e., all the descendants of its single basal species are not included); rather, they are a paraphyletic one, and so from a cla-distic standpoint, should be recognized as such. The situation, though, is complicated because most languages, lack a vernacular term for all the sawflies exclusive of the wasp-waisted taxa. Thus, use of informal terms, typically indicated as such by the use of quotation marks, is seen more and more. Terms like “Symphyta” indicate that the name is being used as a handle of convenience and should not be taken as a reference to a monophyletic group.
The wasp-waisted Hymenoptera have also been traditionally classified into two groups, the “Aculeata” and the “Parasitica,” the former group being largely (though not entirely) distinguished from the latter by the loss of use of the ovipositor for egg laying, being used instead for stinging prey and/or potential enemies. As with the “Symphyta,” the “Parasitica” are no longer recognized as a formal group because there is a reasonable consensus that the aculeates are derived from within them, so rendering them paraphyletic. The Aculeata are strongly supported as monophyletic, at least on morphological grounds.
Sawflies and Wood Wasps
The basal representatives of the order from within which the parasitic and social families of wasps have evolved are the sawflies and wood wasps. There is little doubt that the most ancient extant family of the Hymenoptera is the Xyelidae, represented today by two subfamilies, one whose larvae feed on gymnosperms (either pollen of male cones or on buds or young shoots), and the other having rather caterpillar-like larvae that feed exposed on the leaves of elms and walnut species. Xyelid sawflies possess distinctive antennae (Fig. 3) with the first section of the flagellum very enlarged (possibly as a result of the fusion of multiple segments). Fossil xyelids date from the middle or late Triassic, some 200 mya, and are the earliest known fossil Hymenoptera (Tables I and II ).
Three other superfamilies of typical sawflies are recognized: the Pamphiloidea, which are relatively uncommon (though they include
FIGURE 3 Angaridyela vitimica, a typical xyelid sawfly from the Lower Cretaceous of Baissa in Siberia (body 10.5 mm long). Note the enlarged third antennal segment.
TABLE I |
|||
Most Recent Classification of the Sawflies and Wood Wasps |
|||
| Superfamily | Family | Described extant species | Notes |
| Xyeloidea | Xyelidae | 50 | Most ancient family, with Holarctic distribution |
| Pamphilioidea | Pamphiliidae | 250 | Sometimes called webspinning sawflies after the habit of early instars; some are pest species |
| Megalodontesidae | 40 | Rare group with little known about biology; some feed on Apiaceae and on Rutaceae | |
| Tenthredinoidea | Argidae | 800 | A common group |
| Blasticotommidae | -10 | Usually uncommon, larvae live in a ball of foam of ferns | |
| Cimbicidae | 130 | Occasionally common, often rather large and beelike sawflies | |
| Pergidae | 500 | Principally southern group, especially in Australia and South America | |
| Diprionidae | 90 | Pine sawflies | |
| Tenthredinidae | 4,000 | Very common and speciose in temperate areas, uncommon but moderately diverse in the tropics; most are exophytic with caterpillar-like larvae; some are gall formers | |
| Cephoidea | Cephidae | 80 | Stem sawflies, elongate, associated with grasses and rosaceous shrubs |
| Anaxyeloidea | Anaxyelidae | 1 | A single species from western United States associated with fire-damaged trees |
| Siricoidea | Siricidae | 95 | Horntail wood wasps |
| Xiphydroidea | Xiphydriidae | 100 | Horntail wood wasps |
| Orussoidea | Orussidae | 75 | Parasitic sawflies |
TABLE IIGenerally Accepted Classification of the Apocritaa |
|||
| Superfamily | Family | Described extant species | Notes |
| Stephanoidea | Stephanidae | 200 | No common name; idiobiont parasitoids of wood-boring beetle larvae |
| Megalyroidea | Megalyridae | 50 | No common name; usually rare, presumed idiobiont parasitoids of wood-boring beetle and aculeate wasp larvae |
| Trigonaloidea | Trigonalidae | 100 | No common name; complex hyperparasitic (rarely primary parasitic) life cycle |
| Evanioidea | Evaniidae Gasteruptiidae Aulacidae |
500 420 200 | Ensign wasps; often common in the tropics, endoparasitic or egg-predatory in cockroach oothecae Moderately common, cosmopolitan, kleptoparasites of solitary bees; their larvae kill the host bee’s egg and consume the latter’s pollen food store Koinobiont endoparasitoid of wood wasps and of wood-boring Coleoptera larvae |
| Ceraphronoidea | Ceraphronidae Megaspilidae | 350 450 | Small to very small, very common wasps with a wide range of parasitic biologies Small to very small, common wasps with a wide range of parasitic biologies |
| Platygastroidea | Platygastridae Scelionidae | 1,100 3,000 | Extremely common, usually small, cosmopolitan, mainly koinobiont endoparasi-toids of Diptera larvae, but other biologies and hosts known Extremely common, usually small, cosmopolitan, idiobiont egg parasitoids of many groups of insects and spiders |
| Proctotrupoidea | Austroniidae Diapriidae Heloridae Maamingidae |
3 2,300 7 2 |
Extremely rare, Australian, biology unknown Very common, idiobiont and koinobiont endoparasitoids, mainly of Diptera larvae/pupae Usually uncommon, koinobiont endoparasitoids of Neuroptera (Chrysopidae) larvae Most recently described family, known only from New Zealand, biology unknown |
TABLE II(Continued) |
|||
| Monomachiidae | 20 | Parasitoids of Diptera (Stratiomyidae) in Australia and South America | |
| Pelecinidae | 3 | Moderately common, large, entirely New World, koinobiont endoparasitoids of | |
| subterranean Coleoptera larvae | |||
| Peradeniidae | 2 | Very uncommon, Australian, biology unknown | |
| Proctotrupidae | 310 | Common, mainly northern, koinobiont endoparasitoids, mainly of Coleoptera | |
| larvae, mostly in soil or litter layer | |||
| Renyxidae | 2 | Holarctic, extremely rare, biology unknown | |
| Roproniidae | 18 | Usually uncommon, Holarctic and Oriental, parasitoids of sawflies | |
| Vanhorniidae | 5 | Generally uncommon, New World parasitoids of eucnemid beetle larvae | |
| Mymarommatoidea | Mymarommatidae | 14 | Very uncommon, minute, biology unknown but guessed to be egg parasitoids |
| Chalcidoidea | —20 families | 19,000 | Chalcids; approximately 20 families are recognized, most with diverse |
| biologies | |||
| Cynipoidea | Ibaliidae | 50 | Egg-larval, koinobiont endoparasitoids of wood wasps |
| Liopteridae | 50 | Endoparasitoids (probably koinobiont) of wood-boring Coleoptera larvae in the | |
| tropics | |||
| Cynipidae | 1,000 | True gall wasps and also inquilines in other cynipid galls | |
| Figitidae | 1,500 | Very common endoparasitoids of Diptera, of Neuroptera, and of | |
| hymenopterous parasitoids of aphids | |||
| Austrocynipidae | 1 | Extremely rare, parasitic on Lepidoptera larvae in Araucaria cones | |
| in Australia | |||
| Ichneumonoidea | Ichneumonidae | 22,000 | Very common, biologically diverse though not including any egg parasitoids |
| Braconidae | 20,000 | Very common, biologically diverse though not including any egg parasitoids | |
| Chrysidoidea | Bethylidae | 2,000 | Common and widespread, small ectoparasitoids of small beetle and moth larvae |
| in semicryptic locations; some show parental care | |||
| Chrysididae | 3,000 | Common, cosmopolitan; chrysidines are mainly larval parasitoids of solitary | |
| vespid wasps and bees; cleptines attack sawfly prepupae; others are idiobiont | |||
| egg parasitoids of stick insects | |||
| Dryinidae | 950 | Common, koinobiont parasitoids of larger Auchenorrhyncha | |
| (e.g., Cicadelloidea), some developing partially externally | |||
| Embolemidae | 16 | Uncommon, cosmopolitan; one species parasitic on Heteroptera nymphs | |
| Plumariidae | 20 | Very rare tropical wasps, biology unknown | |
| Sclerogibbidae | 10 | Very rare, koinobiont ectoparasitoids of webspinners (Embioptera) | |
| Scolebythidae | 3 | Very rare, tropical ectoparasitoids (probably idiobiont) of wood-boring beetle | |
| larvae | |||
| Vespoidea | Bradynobaenidae | 200 | Very rare, cosmopolitan, may be koinobiont ectoparasitoids of sun-spiders |
| (Solpugida); great sexual dimorphism | |||
| Formicidae | 10,000 | Ants; extremely common and cosmopolitan; most are eusocial but also includes | |
| social parasites of other ants and slave makers | |||
| Mutillidae | 5,000 | Velvet ants; not true ants, commonest in arid tropics, these are idiobiont | |
| ectoparasitoids of aculeate larvae and pupae in their cells; great sexual | |||
| dimorphism | |||
| Pompilidae | 4,000 | Spider wasps; sometimes large; some are ectoparasitoids of spiders in situ; | |
| most relocate spider prey to a new site | |||
| Rhopalosommatidae | 35 | Moderately common, ectoparasitoids of crickets (Gryllidae) | |
| Sapygidae | 80 | Kleptoparasitoids of solitary bees; usually uncommon but occasionally | |
| a pest | |||
| Scoliidae | 300 | Often large, primarily tropical or warm temperate, idiobiont ectoparasitoids | |
| of subterranean beetle larvae | |||
| Sierolomorphidae | 10 | Very rare, Americas and Oriental region, biology unknown | |
| Tiphiidae | 1,500 | Often common, mainly tropical, idiobiont ectoparasitoids mainly of | |
| subterranean beetle larvae; great sexual dimorphism | |||
| Vespidae | 4,000 | Very common, includes the familiar social wasps or yellow jackets, mason | |
| or potter wasps; females progressively provision their larvae with chewed | |||
| insect/spider tissue, or in Masarinae, pollen | |||
| Apoidea | Apidae | 30,000 | Bees, from solitary to highly social; progressively provision larvae |
| with pollen | |||
| Sphecidae | 8,000 | Almost entirely solitary, nest-building predators of insects and spiders | |
aThe Chrysidoidea, Vespoidea, and Apoidea comprise the aculeate Hymenoptera. The true number of species in the Ceraphronoidea, Platygastroidea, Diapriidae, Chalcidoidea, and Ichneumonoidea, in particular, are likely to greatly exceed the figures given here.
a few pest species), the Cephoidea or stem sawflies, which are rather slender and feed endophytically in grasses and a few woody plant stems, and the very large and speciose Tenthredinoidea. All these are predominantly Holarctic in distribution, although it is likely that at least the tenthredinoids have moderate species diversity in the tropics, even though they are typically quite uncommon there. Some tenthredinoids form galls (e.g., the genera Euura and Pontania), and it appears that secretions from the female sawfly contain the cecidogenic compounds. A few are interesting because they show a degree of parental care (Fig. 4), and in some Australian pergids, the caterpillar-like larvae form resting aggregations during the day but disperse over the food plant at night to feed on the leaves, apparently communicating by means of vibrations.
FIGURE 4 A female pergid sawfly guarding her brood of first instars.
The wood wasps were recognized as a group of three superfamilies, the Siricoidea, Xiphydroidea, and Anaxyeloidea (by Vilhelmsen in 2001), although the greater part of the literature on wasps refers to them simply as Siricoidea. This division is intended to reflect better the phylogeny of the group (Fig. 5), which forms a grade rather than a monophyletic group. Their endoxylous larvae are typically associated with fungus-infected wood, and the adults are responsible for transporting these fungi to their host trees in a special mycangial pouch that is associated with the reproductive system and lies near the base of the ovipositor. In addition to these fungi, at least in some siricid species, components of the secretions injected into the tree at the time of oviposition are actually phytotoxic and may be important in helping their fungi overcome the tree’s defenses. Some siricids are important pests of conifer plantations.
There is a very large body of evidence that the Orussidae form the sister group of all the wasp-waisted or apocritan Hymenoptera and, therefore, its biology is of such interest that the family warrants some separate discussion, small though it is. There have been few detailed biological studies on the group, but it is clear that some, probably all, are parasitic on wood-boring insect larvae, although in at least one species the final instar continues to feed on and in its host when it is fairly well decayed. The orussids cannot be considered to represent the ancestral condition of the parasitic wasps as
FIGURE 5 Working scheme of Hymenoptera higher level phyl-ogeny. The relationships among the basal, principally phytophagous sawflies and wood wasps are becoming well established, but only a few broad groupings are widely accepted among the Apocrita.
a whole because they show many very derived features not found in any other Hymenoptera. It is tempting, however, to consider at least that their biology is closely similar to and perhaps not modified much from the earliest parasitic wasps. Recent studies have shown that they use vibrational sounding (i.e., echolocation through a solid substrate) to detect their host boring. The females’ antennal apices are massive and solid and are used to tap the substrate, and their foretibiae contain massive subgenual organs that are used to detect the vibrations that are transmitted through the wood. This form of host location has evolved on several independent occasions within the order, but the Orussidae are particularly interesting because the egg may be laid into a boring some distance from the host (in the studied species a beetle pupa), and it is the first instar orussid that seeks out the host.
Apocrita or Wasp-Waisted Hymenoptera
In terms of numbers of species, the Apocrita is dominated by parasitoids, but the very great biological developments leading to the evolution of sociality in bees, yellowjacket wasps, and ants has had the effect of polarizing study and discussion, and, in the past, also classification. It has long been appreciated that the aculeate Hymenoptera, which include the social taxa, have evolved from within the Apocrita (in fact, they seem to be most closely related to one particular superfamily, the Ichneumonoidea).
THE “PARASITICA”
The “Parasitica” is a paraphyletic group (with respect to the aculeates) that comprises some 11 or 12 super-families, the exact number being a little unstable because rapid advances in phylogenetic studies are tending to suggest that some previously recognized taxa are paraphyletic or polyphyletic. Consequently, the number is likely to increase by a few when new evidence leads to robust and better resolved phylogenetic trees. A rather conservative phylogenetic hypothesis is shown in Fig. 5, but even this is problematic. Areas of particular uncertainty are the monophyly of the Proctotrupoidea (even after the Platygastroidea, once included therein, have been removed to a separate group), and of the evaniomorph superfamilies
(viz., Megalyroidea, Evanioidea, Trigonaloidea, Ceraphronoidea, and Stephanoidea). Even within this grouping it is not yet totally certain that the three families constituting the Evanioidea form a mono-phyletic group; they have very different biologies, and no unique synapomorphies have been found to unite the component families.
The vast majority of known and undescribed species are parasi-toids of other insects. The term “parasitoid” is used almost universally nowadays to distinguish the interactions of these insects from those of parasites, although purely for reasons of euphony, we still often refer to them as parasitic wasps rather than parasitoid wasps. Whereas true parasites live off the living bodies of their hosts, they seldom kill them; indeed, it would usually be maladaptive for them to do so because the longer the host lives, generally the greater will be the reproductive opportunity of the parasite. Parasitoids, however, always kill their host and treat it as a single meal that will provide all the food necessary for their own development. Once the parasitoid wasp has eaten all of the host that it needs, the host is no longer of use, and so there is no need to leave it to recover. In this respect, parasitoids are rather akin to predators but, unlike predators, they require only a single host (i.e., prey) individual to provide all their needs. (Predators, on the other hand, eat multiple, often very many, prey during their life span.) Even so, a few groups of “parasitic” wasps actually behave more like predators. These include species that attack egg masses of, for example, spiders, and species whose larvae will eat not one egg but many or all in the spider’s batch.
IDIOBIONTS AND KOINOBIONTS
Almost all parasitoids can be classified into one of two classes defined by whether their hosts continue developing after parasitization (the koinobiont strategy) or whether further host development is curtailed at that time (the idiobiont strategy). These strategies explain many other aspects of the wasp’s biology, such as longevity, egg development, egg size, and fecundity, and host range, as discussed in greater detail elsewhere in this volume and in works by Godfray and Quicke. There has been general agreement that the first parasitic wasps were probably idiobionts. However, it is far from proven that this means that within any given parasitic wasp family or superfamily, the idiobiont taxa are necessarily ancestral. Indeed, the most parsimonious explanation for the distribution of parasitoid life history strategies on several of the most widely cited phylogenies suggests that endopara-sitoids have evolved into ectoparasitoids on many occasions within the Chalcidoidea, though probably not within the Ichneumonoidea.
Aculeate Hymenoptera
The very familiar, often social, ants (Formicidae), bees (Apidae or Apiinae depending on classification system), and yellowjacket wasps (Vespidae) have rather dominated the traditional classifications of Hymenoptera. These taxa are united with a number of other, rather less familiar ones to form a monophyletic group called the Aculeata. Most members of this clade have derived biologies, though some families are still functionally parasitoids. In these, however, the ovipositor proper is not used for reaching hosts, and only in a few (possibly derived) cases is it actually used for passing the egg. Even among the essentially parasitic species there is a strong tendency toward physically manipulating hosts. Many spider wasps (Pompilidae), for example, drag paralyzed hosts to a hiding place before ovipositing on them, sometimes biting off the spiders’ legs to facilitate handling and carrying. The least-derived biology is to use a preexisting cavity as a hiding place, and subsequent evolution has led to many levels of modification or de novo nest construction.
SOCIAL KLEPTOPARASITES AND SLAVE MAKING
The great resources afforded by social insect nests have attracted the evolution of numerous thieves, some of which are other hymenopterans, often closely related to the species being robbed. The most interesting of these interactions are exhibited by social kleptoparasites when a female of a nonsocial species replaces the queen of a social one and makes use of the workers in the host colony to rear her own brood, rather than relying on more of their own siblings. This ability obviously depends on sophisticated mimicry, not just visual, but more particularly chemical and tactile, or otherwise the usurped workers would detect the intruder. Social kleptoparasitic taxa have evolved independently on numerous occasions and are known among bees (Apidae), vespid wasps, and ants. “Slave making,” another form of social parasitism, is known only among the ants, but within this family it has been evolved by several different lineages. The workers of one species are abducted by the slave-making species and forced to forage for the latter.
RELATIONSHIPS TO MAN
Beneficial Species
The number of beneficial hymenopterans greatly outweighs the number of harmful ones, though some taxa fall into both categories depending on circumstance. Even the vespid wasps (yellowjackets), which are well-known nuisances at picnics and barbecues, and occasionally have serious medical consequences, are responsible for eating a very large number of other insects (their larvae are fed almost entirely on chewed-up insect muscle), and in agricultural settings undoubtedly devour many pest insects. Humans have made use of the voracious insect-eating capacities of some ants for many years. For example, Chinese citrus growers have traditionally transferred tree ants, Oecophylla spp., into their orange groves to consume potential pests, and central European foresters have had considerable success in controlling pest outbreaks by transporting into forests the nests of wood ants, Formica rufa.
Biological control programs have made a great deal of use of parasitic Hymenoptera to control host pest populations, sometimes with spectacular results, and when such controls work, the cost-benefit ratio is very favorable. It is becoming increasingly possible to use commercially produced parasitic wasps to control pests in private gardens and greenhouses as well as on large commercial enterprises, and this has obvious desirable features such as reducing the need to apply pesticides.
Bees and Honey
Probably best known of the beneficial affects of Hymenoptera is the production of honey, which people have valued as a sweetener since prehistory. Originally various honey-producing wild bee nests would have been harvested, and indeed are still harvested by indigenous peoples on several continents. However, one species, the European hive bee or honey bee, Apis mellifera, was found to be both productive and manageable, and it was effectively domesticated by getting it to nest in artificial hives several thousand years ago. The first records of beekeeping come from ancient Egyptian wall paintings, some 2500 years b . c. Bees also produce wax from which their larval cells are formed, and this is also widely used for a variety of purposes, from candles to cosmetics and from pharmaceuticals to polishes.
Although most people think of A. mellifera when they talk of bees, there are in fact more than 25,000 bee species in the world, the majority of which are solitary rather than social. Collectively, bees are of immense economic importance because of their major role in plant pollination, which includes the pollination of a large number of crop species, and this service is far more valuable than the production of honey and beeswax. In addition to the economic importance of bees, most ecosystems rely on them for pollination, largely because bees have evolved, diversified, and specialized together with the angiosperm plants they pollinate, and it cannot be doubted that loss of many bee species would have a long-term dramatic impact on the floral composition of many habitats. Interest has been increasing of late in non-Apis bees because of the realization that they may be excellent pollinators of many crops, sometimes better than honey bees, and that changes in habitat use have led to a decline in many of these useful species. Additionally, Varroa, a parasitic mite that has spread from its natural eastern honey bee host, A. cerana, onto the common honey bee, has resulted in serious losses of the latter and thus in pollination rates.
All bees feed their larvae on pollen that is harvested and processed by the adult females (worker bees among the highly eusocial taxa). Typically the pollen is mixed with nectar, although plant oils are sometimes used. The honey made, for example, by Apis is not a food for the bee larvae at all, but rather a high-energy food source that is produced and stored by the social bees for times when their normal source of energy food, nectar, is in short supply or simply unavailable—as in temperate winters.
Pests and Medical Importance
Considering the great abundance and species diversity of the Hymenoptera, it could easily be argued that the order includes very few pests. There is general awareness that many bees and most social wasps are capable of delivering a quite painful sting, but for most people these stings are relatively minor, if unpleasant, quickly forgotten occurrences. In the United States approximately 50 people die each year as a result of hymenopteran stings. To put this statistic into perspective, about 25 people are killed each year by lightning in the United States.
The major cause of death from hymenopteran sting is respiratory block, sometimes as a result of a sting in the inside the throat as a consequence of inhaling a wasp or bee, but also often because of more systemic reactions. Approximately 2% of the population is hypersensitive to Hymenoptera stings, but although such sensitiza-tion puts the victims at a greater risk, only about one death per year in the United States is attributable to a hyperallergenic response.
Some sawflies and gall wasps are economically important pests of crop and ornamental plants, but many of these have potential roles in the biological control of weeds as well. Notable pest saw-flies include both external foliage feeders such as the pine sawflies (Diprionidae), which can be major defoliators of coniferous forests, and concealed feeders such as the wheat stem sawfly, Cephus cinctus (Cephidae) and the siricid wood wasp, Sirex noctilio, with the latter having caused considerable harm to pine (Pinus radiata) plantations in Australia and New Zealand.
Probably the largest number of pest taxa is to be found among the ants. There are a number of introduced taxa in many parts of the world that cause considerable damage to crops, are harmful to livestock, and have nuisance value to humans. Three of many possible examples will serve to illustrate the range of harmful interactions that can occur. The tiny pharaoh ant, Monomorium pharaonis, a tropical species widely introduced into buildings in the Northern Hemisphere, often gets into hospitals, where it is a potential vector of bacteria—it has the habit of getting into almost anything in its foraging, including under bandages. The Argentine ant, Linepithema humile, is a highly invasive species that fights and outcompetes other ants in introduced regions, and its huge colonies cause damage in agricultural situations because they protect aphids and other honeydew-forming pest insects. Third, the red imported fire ant, Solenopsis invicta, which has spread alarmingly in the southern United States in recent years, is very aggressive, especially when its mounds are disturbed. Fire ants will sting people, pets, and livestock many times, with resulting pain, blisters, and even systemic reactions, and small pets are sometimes killed by them. Interestingly, many of these pests are far less harmful in their native regions.
IDENTIFICATION
As might be expected with such a large insect order, identification of hymenopteran species, but also genera and even families, is not without its difficulties. Even in parts of the world where the insect fauna is quite well known, such as Europe and North America, there will be many groups for which there are no satisfactory identification keys. In less well-known parts of the world, almost any reasonably sized sample will contain numerous undescribed species and within some families even genera. The presence of so many unclassified hymenopterans reflects a combination of innate taxonomic difficulties such as the insects’ small size, the large number of similar species, and often a great deal of superficial convergence. In addition, because the order, with few exceptions, has not attracted a great deal of amateur attention, relatively little work has been done on it. Recent years have, however, seen vast improvements to the situation. Several well-illustrated keys to all families have been published, as well as several major works on the more popular aculeates. Particularly useful works are by Gauld and Bolton, and Goulet and Huber.
